On Wed, 25 Oct 2023 09:46:13 +0100, Oliver Upton <oliver.upton@xxxxxxxxx> wrote: > > To your point, the SPSR_* accessors still trap even if AArch32 is not > implemented. I was suggesting in passing that it'd be nice if the > architecture alternatively allowed for these to read as RES0 (no trap) > if NV==1 and AArch32 is not implemented, which aligns with your change. Ah, I see what you mean now. Yeah, I'll put that on the bucket list of things I need to write up. It has quite a few entries already, and I really should get to it. Thanks, M. -- Without deviation from the norm, progress is not possible.
